Defense Technical Information Center Compilation Part Notice

Transcription

1 Defense Technical Information Center Compilation Part Notice ADPO TITLE: Wind Tunnel Model Using and Testing Using Rapid Prototype Materials and Processes DISTRIBUTION: Approved for public release, distribution unlimited This paper is part of the following report: TITLE: The Annual AIAA/BMDO Technology Conference [10th] Held in Williamsburg, Virginia on July 23-26, Volume 1. Unclassified Proceedings To order the complete compilation report, use: ADB The component part is provided here to allow users access to individually authored sections f proceedings, annals, symposia, etc. However, the component should be considered within [he context of the overall compilation report and not as a stand-alone technical report. The following component part numbers comprise the compilation report: ADPO11183 thru ADPO11193 ADP thru ADP UNCLASSIFIED

2 WIND TUNNEL MODEL DESIGN AND TEST USING RAPID PROTOTYPE MATERIALS AND PROCESSES Richard R. Heisler and Clifford L. Ratliff The Johns Hopkins University Applied Physic Laboratory Laurel, MD Abstract Reduce model fabrication costs by more than 50% of a traditional steel and aluminum model. Whether an airframe is a new design, modification Zý of an existing design, or evaluation of a competing or Program Accomplishments foreign design, an accurate, high-confidence representa- RPM-01 was designed and fabricated at JHU/APL and tion of the airframe aerodynamics is paramount to any tested in the Lockheed Martin Missile and Fire low-risk design or evaluation effort. These aerodynamic Conte D in the 20kh2 A ugus t and The estimates are used for vehicle and component sizing, Control-Dallas HSWT from August The performance estimates, and autopilot design and evalu- model was built in less than half the time and for ation. The advent of new rapid prototyping manu- approximately one-third the cost of WTM-01. Fiftyfacturing techniques and materials could provide a three tunnel data occupancy. runs were obtained during hours of Data were obtained for angles of means to reduce the cost associated with the acquisition of a wind tunnel model (WTM), provided the data attack (a) up to 20' and sideslip angles (P) to 10' for obtained with the rapid prototype model (RPM) were of Mach 0.40 to Various configurations and fin and sufficient fidelity to justify its use. The Johns Hopkins canard materials were tested. A complete run log is University Applied Physics Laboratory (JHU/APL) has provided in Figure 1. developed a WTM design that consists of a steel The results of the testing demonstrated that the hybrid strongback with rapid prototype plastic components WTM (ABS fuselage with a steel strongback core) was attached to it to provide the overall vehicle configuration. In a test at the Lockheed Martin Missile not only more cost effective than a similar all-metal model, but provided acceptable data fidelity as well. and Fire Control-Dallas High-Speed Wind Tunnel Extensive post-test analyses determined the cause for (HSWT), this model (designated RPM-01) demon- failure of some components and identified candidate strated acceptable data quality and test-to-test data epoxies with stronger material properties and temperarepeatability with a geometrically similar steel and ture resilience. aluminum model. RPM-01 was built for approximately a quarter of the cost and a fraction of the time required Model Description of the all-metal high-fidelity model. Program Obiectives In 1997, JHU/APL built a high-fidelity 25% scale steel and aluminum model of a canard wing airframe with an underbody flow through inlet and integral top-mounted The objectives of this research program were to demonstrate the feasibility of using a fused deposition launch lugs. This model, designated WTM-01, would provide the cost, schedule, and data quality benchmarks modeling (FDM) process to create acrylonitrile! for RPM-01, which was fabricated for this research butadiene/styrene (ABS) plastic parts of suitable effort. The WTM-01 model canards and tails are strength and fidelity to build a WTM in a fraction of the arranged in the + and x orientation, respectively, when time and cost associated with a traditional all-metal the vehicle is at 00 roll attitude. The horizontal canards WTM. The measures of success associated with this are fixed at 50 leading edge up while the vertical program were to canards are deflectable to provide yaw control. The four "* Demonstrate the model's ability to survive the tails incorporate trailing edge elevons that deflect to wind tunnel test environment, provide airframe pitch and roll control. An inlet fairing for the model was designed to prevent flow through the "* Deliver data quality that is within JHU/APL's inlet and present a clean aerodynamic surface to the previously achieved test-to-test repeatability levels. oncoming flow. UNCLASSIFIED DISTRIBUTION STATEMENT A: Approved for public release; distribution is unlimited. 5-11

3 Mach Configu- No. RN/ft ration =0'O 90' at = 0' at = 5' a= 10 a = 20' = 0 0 = ' M 1 11 (B) 12 (B) 13 (A) 14 (A) 15 (A) 2 31 (B) 32 (B) 3 51 (B) 52 (B) 4 71 (C) 66 (C) 65 (C) 64 (C) 61, 62 (B) 63 (B) 181 (B) (B) 102 (B) (B) 122 (B) (B) 142 (B) M 4 82 (C) 86 (C) 85 (C) 81 (B) 83 (B) 84 (B) M 1 21 (B) 22 (B) 23 (A) 24 (A) 25 (A) 2 41 (B) 42 (B) 4 92 (C) 96 (C) 95 (C) 91, 93 (B) 94 (B) 171 (B) (B) 112(B) (B) 132 (B) (B) 152 (B) Configuration Codes: 1. Body alone 2. Body + canards (vertical canards: metal, horizontal canards: infused plastic) 3. Body + canards (vertical canards: metal, horizontal canards: cast resin) 4. Full configuration = body + 4 metal canards + 4 metal fins + inlet + lugs 5. Full configuration = body + 4 metal canards + 2 metal fins (1 and 4) + 2 rapid prototype fins (2 and 3) + inlet + lugs 6. Full configuration = body + 4 metal canards + 2 metal fins (1 and 4) + 2 infused plastic fins (2 and 3) + inlet + lugs 7. Full configuration = body + 4 metal canards + 4 cast resin fins + inlet + lugs Sweep Codes: Roll: A = 0 _<<270' Pitch: B 00 _< 0 < 200 Yaw: C 00<1 3 < 100 Figure 1 HSWT 1307 Run Log The model was tested primarily with the fairing in This design approach was chosen to address place; however, data with the inlet flowing were deficiencies identified in earlier research performed by obtained to assess the aerodynamic impact because of Marshall Space Flight Center (MSFC) (References 3 the flowing inlet. WTM-01 was tested in the Veridian through 5). Their research focused on the various rapid Calspan 8x8-ft transonic wind tunnel in Buffalo, NY, at prototype methods and materials available and which of the following test conditions: these provided the best performance with regard to " Mach: 0.40< M:<0.90 model construction and data quality. Their research concluded that "rapid prototype methods and materials "* Angles of attack: -5' < ca:< 200 can be used in subsonic, transonic, and supersonic "testing for initial baseline aerodynamic database * Angles of sideslip:-10< 3 < 100 development" and "the accuracy of the data is lower A more detailed model description and post-test report than that of a metal model due to surface finish and are provided in References I and 2, tolerances." Their research indicated that the models made using the FDM-ABS and stereolithography (.stl) RPM-01 is a 15% scale model of the above airframe, materials and processes provided the best results. consisting of a steel strongback or backbone with ABS However, the FDM-ABS model data did diverge from plastic parts attached to it to create the outer moldline the metal model results at high loading conditions, thus of the model. The aerodynamic surfaces (canards and producing unsatisfactory results. These differences fins) for the model pass through the ABS plastic skin were attributed to surface finish, structural deflection, and attach directly to the strongback with screws. A and tolerance deviations when the material was grown. schematic diagram of the RPM-01 buildup is provided in Figure 2, and a detailed description of the model The JHU/APL model was fabricated using the FDMdesign and construction is provided in Appendix A. ABS material and rapid prototype manufacturing process. The steel strongback used in the model provides rigidity to the complete model and allows the UNCLASSIFIED 5-12

4 aerodynamic surfaces to be mounted directly to it. This and pertinent dimensions is provided in Figure 3. The reduces or eliminates any structural deflection the machine used to manufacture the plastic components model body and metal fins might experience in the test. has a 9-inch maximum limit on component length and a The ABS plastic and cast resin fins did experience quoted manufacturing tolerance of ±0.005 inch. For flexure and sometimes failure under load during the RPM-01, this tolerance scales to ±0.030 inch on the test. full-scale vehicle, which is comparable to a full-scale Another advantage provided by the strongback design manufacturing specification. If RPM-01 had been is the ability to make larger models and reduce the limited to the maximum 9-inch length as were the uncertainties because of manufacturing tolerances and MSFC models, the same ±0.005-inch manufacturing model scale. RPM-01 was a 15% scale model with an tolerance becomes ±0.094 inch full scale, well in excess overall length of inches. A sketch of the model of the full-scale specification. Reusable Steel Strongback ABS Plastic Faired Inlet Facility- Provided Balance ABS Plastic Nose, FuselageBanc BoattailBanc Adapter Figure 2 Model Exploded View LIIIII Steel components Plastic, fused deposition components Steel, injection cast resin, or fused deposition fins and canards Reusable Steel Backbone Balance Center Facility-Provided Balance M.S. Figure 3 Model Schematic UNCLASSIFIED 5-13

5 RPM-01 obtains its size by employing three cylindrical Test Facility interlocking components that fit tightly over the strongback to form the fuselage/boattail assembly. The The Lockheed Martin Missile and Fire Control-Dallas model nose attaches separately to the strongback via a HSWT is an intermittent blow-down-to-atmosphere threaded spindle assembly. The model inlet, fins and tunnel with a Mach number capability of 0.30 to 4.80 canards passed through the plastic fuselage skin to and dynamic pressure range from 300 to 5000 psf. The mount directly to the strongback for strength. The tunnel is operated by the controlled discharge of model launch lugs attach to the plastic skin with screw compressed air supplied by eight storage tanks that are and helicoil attachments. To test a clean symmetric charged by three series-connected, multistage, centrifucone-cylinder configuration, rapid prototype filler gal compressors driven by an 8000-hp electric motor. blocks for the canard, fin, and inlet access holes were Air storage volume is 40,000 ft 3 with a maximum manufactured. This allowed the model's symmetry storage pressure of 500 psi at 100 F. characteristics angle. to be quantified as a function of roll The tunnel uses both supersonic and transonic test sections, depending on the desired test conditions. Each Aerodynamic Surfaces test section is 4x4 ft in cross-section and approximately 5 ft long. For operation at Mach >1.50, the supersonic sof missile control surfaces (i.e., canards and test section is employed. A single-peak variable Fou ersets diffuser fis located downstream of the test section and allows tails) were fabricated for testing: steel, ABS plastic,s ABS plastic with vacuum epoxy infiltration (vacuum- w pressures than would be infused ABS), and cast mold C1511 urethane resin. The possible with a fixed diffuser. For subsonic and tranuse of the steel tails and canards provided a rigid sonic operation (0.30 < M _< 1.50), the variable diffuser t' is removed and replaced with the perforated wall reference for comparison against the nonmetallic tails isaremoved a eced with the perorated wal to discern elastic and nonlinear behavior as a function transonic test section having 22% porosity. The of aerodynamic loading in subsonic Zý and transonic Mach transonic main plenum is pumped by ejector action of the tunnel airstreams acting on controllable ejector number regimes at various model roll and pitch main tunnel airstream actin on controll able orientations. flaps located downstream of the test region. Adjustable choking flaps, also located downstream of the test The ABS plastic surfaces were created (grown) on the region, are used to control Mach number. FDM machine whereby model parts are built up in thin slices or roads (i.e., inch) of ABS plastic with The HSWT sting cart is a servo-controlled hydraualternating pattern directions. Even though the parts lically actuated cart capable of a -13' to +23' sweep were manufactured using the maximum density range while maintaining the center of rotation on the possible, the resultant components are still porous. As tunnel centerline. Fixed offset adapters can be installed an inexpensive method to fill the microscopic pores and to extend the available pitch range. Sweep rates up to 5' increase stiffness, the ABS components were immersed per second are available. A remotely controlled roll in an epoxy bath while under vacuum pressure. This sting can be added to provide multiple pitch, roll, and process, deemed "vacuum infusion," results in a more yaw sweep capability. solid component because air in the porous areas has been displaced by the epoxy. Test Results The third nonmetallic surfaces tested were cast resin As shown in Figure 1, the RPM-01 test was conducted molded surfaces. These were manufactured by creating in three phases. Phase I consisted of the RPM-01 body a mold of a specially prepared ABS part that was grown alone. This was necessary to assess any model or tunnel on the FDM machine. The surfaces of the part are asymmetries that may be present in the data and to filled, sanded, and primed to provide the smoothest provide a baseline for the alternate nonmetallic canard possible surface finish and then treated with a silicon studies to be done in phase 2. Both roll sweep data at release agent. This finished part is then used as a master fixed pitch angles and pitch sweep data at fixed roll to create a silicon mold. Vent holes are placed in the angles were obtained. Phase 1 test results (Figure 4) mold, and the desired epoxy resin is injected into the indicate that good data symmetry was obtained between mold cavity using a hypodermic needle. The aero- the quadrants. Because four interlocking cylindrical dynamic test results obtained with these components components were required to build up the fuselage, any are discussed later in this paper. tolerance problems or mismatches created when the parts were grown would have been apparent in this data. Therefore, the ABS plastic cylinders built up over the steel backbone possess sufficient fidelity (and UNCLASSIFIED 5-14

6 efficient means of gathering data for both trim and non- trim orientations. roundness) to allow acceptable roll sweep data to be obtained. The ability to obtain roll sweep data is important in missile testing because it provides an 2 - SMach = Mach = " _ I Phi - Degrees Figure 4 Phase 1 (Body Alone) Test Results The alternate canard material studies conducted during The spindle was the point of failure for the vacuumphase 2 were done on the body/canard configuration. fused ABS canards. A better approach would have been For these runs, undeflected steel canards were installed to use a unique design for the ABS canards that would horizontally, and the alternate material canards (ABS maximize the cross-section at the intersection with the plastic, vacuum-infused ABS plastic, or resin cast) were component base. installed vertically. Pitch sweep data were obtained at Finally, the failure of the cast resin canards was very 0' and 900 of roll. Increments between the phase I and disappointing and unexpected because temperature was phase 2 runs were taken to assess the fidelity of the data acquireda factor considered when the casting resin was chosen. It was determined that for an airfoil application such as The results of the body-canard testing are provided in this, the thin wafer resin properties need to be evaluated Figure 5. It can be seen that at Mach 0.40, the vacuum- rather than the more conventional cylindrical infused ABS canard results are comparable to those properties. Several alternate resins were evaluated obtained with the steel canards. The resin cast canards following the wind tunnel test to assess their were unable to withstand the air temperature during the temperature and strength properties. These results are run and became extremely pliable under load, resulting discussed further in Post-Test Failure Analysis and in unacceptable performance at Mach They were Testing. not tested agyain at Mach An indication of the Phase 3 was the full-up configuration comparison to the achieved in-test repeatability at Mach 0.40 is also high-fidelit WTM-01 data. Pitch swee runs at 00 and provided in Figure 5. At Mach 0.90, the vacuuminfused canards provided acceptable results up to sideslip and yaw sweep runs at 0', 100, and 20' angle of attack. At that point, one canard failed angle of attack were done. Steel canards were used for completely, resulting in the reduced normal force all the comparisons, but steel and alternate material fins coefficient shown in Figure 5 and a subsequent rolling were weetse.fgrs6 tested. Figures 6 througrh thruh8so 8 show that haodts-o good test-to- momentcoefficien shown tnot ifitest. and facility-to-facility repeatability was obtained moment coefficient not shown. between WTM-0I and RPM-01 with steel fins. Figure 9 To save money during the RPM-01 model design and fabrication, the same design was used for both the metal and non-metal canards. The design used a spindle that attached to a base that allowed the canards to deflect. includes facility-to-facility uncertainties overlaid on the data. These uncertainties were derived from historical Standard Missile test results using high-fidelity steel models in different facilities. These results quanti- UNCLASSIFIED 5-15

7 tatively demonstrate that the RPM with steel fins and plastic fins. Once again the resin fins became soft and canards provides acceptable test results when compared pliable during testing, but because the fins were much with a high-fidelity steel model in another facility, thicker than the canards, the resulting bending was less but still unacceptable. The ABS fins tested were Finally, the feasibility of using nonmetal fins was btsiluacpal.teasfn etdwr F y tinstalled in the upper and lower right fin locations with investigated during phase 3. Fins were made using the steel fins installed on the left side of the model. Both same materials and processes described previously for sets of ABS plastic fins survived testing at Mach 0.40 the canards. Complete sets of resin fins were made, as and 0.75 and 0' sideslip testing at Mach were pairs of ABS plastic and vacuum-infused ABS - Infused ABS Canards Caat resin canard -Steel Canards Steel Canards repeat 2 Mach I.- C.Z Alpha - Degrees "--*Infused ABS Canards -Steel Canards 2- Mach z I-I 0 -i Alpha - Degrees Figure 5 Phase 2 (Body/Canard) Test Results UNCLASSIFIED 5-16

8 -*-WTM01 Calspan 5404 ~ RPM01 HSWT AWRPM01 Vacuum Infused Tails RPM01 Cast Resin Tails 5 - f~-~~ # T Alpha - Degree Figure 6 Test-to-Test Comparison, Mach WTM01 Calspan 5404 ~ RPM01 HSWT J Alh e Figure 7 Test-to-Test Comparison, Mach =0.75 UNCLASSIFIED 5-17

9 -'"WTM~01 Calspan 5404 *RPM01 HSWT RPM01 Infused ABS Fins RPM01 Cast Resin Fins j _ Alh ems Figure 8 Test-to-Test Comparison, Mach =0.90 High Fidelity Steel Model...Rapid Prototype Model Note: Error bars indicate levels of previously obtained 0-_ 1 o test -to-test repeatability E 0~ 0.p4 - ere Fiur -9. TettoTs Coprsn Mac 0.9 wit Unetit-vra 0.L SSF2D51

10 During a blow, pitch sweep data up to 200 angle of Post-Test Failure Analysis and Testing attack were obtained at 00 sideslip. The model was then snap rolled air-on at 20' angle of attack to obtain the Following the test, several new RPM-01 resin tails were 100 sideslip orientation, and data were acquired as the cast and statically tested to identify better materials for model was pitched back down to 0' angle of attack. The use in future testing. Because the mold had already ABS fins failed during the model roll to 100 sideslip at been made, creation of new fins was relatively simple z' and inexpensive. The resins for these fins included (x the = snap 200. The roll aerodynamic combined to loading fracture and the dynamics ABS fins due to at the Smooth-On C1511 (rigid urethane mounting casting pad compound), where the aerodynamic surface Dexter@ EL355, Dexter@ Hysol 9396, and Hysol 9396 intersected the base. The fin load at this condition was +18% end-milled fiberglass fill. In addition to the intersti new ted using computional Th flud at ndyiction bs cast fins, a new vacuum-infused ABS tail was made estimated using higher strength epoxy as the filler infused into a 55 pounds normal force. Static testing of the component mo rep s th oai (ResfinlFusionf860d1/86 confirmed the failure under aerodynamic loading, more porous ABS tail (Resin Fusion 8601/8602 Additional results of the post-test failure analysis are infiltrated to inch honeycomb structure). discussed later. A thin wafer test sample of each new tail material also A comparison of the data obtained using the nonmetal was created at the same time to evaluate the temperafins is provided in Figures 6 and 8. It can be seen that at ture sensitivity of the tails as part of a dynamic material low Mach numbers, the ABS plastic fins provide good- analysis; test results are provided in Figure 10. As quality data. However, as the speed and angle of attack shown in the figure, the C1511 resin is sensitive to increase and fin loads increase, data fidelity suffers temperature and suffers an order of magnitude strength because of fin flexure. Based on these results, the loss between 50'C and 75 0 C (122 F to 167 F). The authors conclude that the ABS plastic fins are not tunnel air was approximately 130'F during testing and suitable for high-speed testing, but further research into explains the observed failure during air-on testing. The cast resin materials is necessary to identify a resin or Hysol 9396 epoxy would have been a much better compound that can withstand the test environment, choice for testing at this facility. 1.E+10I 10o C/min, 1 Hz Data 1.E+09,., 'J~ -. i [u 1.E* T Tunnel T Temperature % o X-. *4 X4~-. 1.E A (3--. EL355 * ABS... Infiltrated ABS 1.E Temperature (C) Figure 10 Material Sensitivity to Temperature UNCLASSIFIED 5-19

11 The failure of the ABS tail during testing was not unexpected. The failure of the vacuum-infused tails and the way they failed raised some questions. Following the test, an X-ray analysis of the vacuum-infused tails revealed the vacuum infusion did not permeate the tail as expected. This result (Figure 11) indicates that only about 65% of the tail was reinforced with epoxy and led to development of the 0.05-inch honeycomb structure for static testing. the RPM-01 strongback (Figure 12). The static load cell testing validated that the Hysol 9396 epoxy resin has superior stiffness properties based on the increased slope of the curve presented in Figure 13. The Hysol and EL355 tail surface did not fracture at the root as with the ABS, but actually fractured at the base around the hole pattern where screws were used to mount the tail to the mounting bracket. Only the ABS and C1511 nonmetallic materials were wind tunnel tested. All of the static load ABS tail structures fractured at the root chord near the base similar to what happened during the wind tunnel test. These results are provided in Figure 13. Figure 11 Vacuum Infusion X-Ray Infiltration Results Finally, all tail samples were tested to failure in a load cell mounted in the same manner as the tails mounted to Figure 12 Static Load Test of RPM-01 Tail II 120 Hys., 9396 Resain ' 80 v I11C1512 Resi 0 I Deflection (Inches) Figure 13 Static Load Test Cell Results UNCLASSIFIED 5-20

12 screw holes for attachment of the canards, tails, and faired inlet were located in the strongback to meet model design requirements. Finally, the inside forward The results of this project demonstrated that the use of end estrn was mined to 2.15anc end of the strongback was machined to a inch rapid prototype manufacturing technology and diameter and threaded to accommodate the spindle for materials can significantly reduce the time and cost attachment of the model nose. The spindle is also associated with the fabrication of a scale model suitable fabricated of 304 stainless steel and has a inch for wind tunnel testing. It was further demonstrated that diameter threaded base for mating to the strongback and by using an innovative hybrid design consisting of ABS a 0.5-inch diameter threaded spindle for attachment of plastic parts built on to or attached to a reusable steel the AS plastic nose. strongback, the data obtained with such a model are comparable in quality to data obtained on an all-metal The ABS plastic nose, fuselage components, boattail, high-fidelity model. These capabilities were demon- faired inlet, fins, canards, launch lugs, and assorted strated in the transonic regime (up to Mach 0.90) using filler blocks were all manufactured using a Stratasys metal fins and canards. Post wind tunnel test research Inc. FDM-1650 machine. The parts were designed and and analysis of the nonmetal aerodynamic surfaces solid geometry models were created using Pro-E design verified the in-test failures of these surfaces and software and output as an.stl file. (This is an output identified the materials and manufacturing processes option built into Pro-E.) The geometry information better suited to withstand the wind tunnel test contained within the.stl file was then mathematically environment. broken down into horizontal slices and transferred to The authors conclude that these new surfaces not only the FDM machine for fabrication. could survive supersonic test conditions, but would The FDM-1650 operates at high temperature to melt the have sufficient strength to deliver accurate data under ABS plastic (or wax). These materials are fed into a these conditions as well. A wind tunnel test, to be temperature-controlled extrusion head, where they are conducted in 2001, will hopefully demonstrate the heated to a semi-liquid state. The melted plastic comes ability of these nonmetallic surfaces at supersonic out in an extruded string of hot liquid and paints an conditions. ultra-thin layer of plastic inch thick onto a fixtureless base. The layers are built one on the other. The combination of these manufacturing technologies The material solidifies, laminating the preceding layer. and materials should make wind tunnel testing a viable Because the plastic is hot and therefore very pliable, a and affordable option to programs developing a new configuration coniguaton or r evaluating valatig a prviusl previously utese untested supporting prototype pieces. system is built underneath to support the configuration. The availability of wind tunnel data early in the design process would greatly reduce the risk Because of FDM-1650 size constraints, the fuselage associated with a configuration down select and could was built up as three sections plus the nose. The inner preclude surprises discovered later in the full-scale diameter of the components was chosen to be equal to development phase of a program. the outer diameter of the strongback. Because of shrinkage during fabrication, the inside of the Appendix A - Fabrication Details of RPM-01 components were lightly sanded to achieve a near-zero tolerance fit over the strongback. The outer surfaces of RPM-01 was manufactured using FDM-ABS plastic the components were sanded to smooth the surface and panels attached to a cylindrical steel strongback. The remove the burrs that accumulate as the part is grown strongback provides strength and rigidity to the plastic and then sprayed with an aerosol solvent (Sandfree) and model and allows larger scale models to be built as wiped clean to produce a smooth clean surface. There well. The strongback, fabricated from 304 stainless are no attachments between the ABS fuselage steel, is a inch long cylinder with a 2.25-inch components and the strongback. outer diameter and a inch inner diameter. The Each fuselage component has an interlocking tab to surface of the cylinder has a surface finish of 32. locate and attach it to the next component. Longitudinal The balance adapter was fabricated by the Lockheed and rotational position is maintained on the strongback Martin Missile and Fire Control-Dallas HSWT, which via attachment of the fins, canards, faired inlet, or filler also performed the final honing of the strongback inside blocks for these components. The slots and holes diameter to achieve a zero tolerance fit between the required in the individual fuselage components that balance adapter and the strongback. (Note: The inner allow the fins, canards, and inlet to attach directly to the surface of the strongback was only machined from one strongback are incorporated into the design during the end to accommodate the balance adapter.) The Pro-E geometry development and included in the.stl remainder of the strongback is unfinished. Threaded file description. As such, these geometry details are UNCLASSIFIED

13 created as the part is grown on the FDM The break-away plane. The removal of the rapid prototype screw holes to attach the launch lugs were drilled into tail left an internal cavity of the tail. With the use of the center and aft fuselage sections after fabrication on some vent holes, a hypodermic needle was used to the FDM machine. Helicoils were inserted into the pressure fill the cavity with several types of epoxy drilled holes to provide the threaded surface. resins. If high-pressure injection plastic casting is The nose is created the same as the fuselage required, a hard tool mold can be made from the staged components with the exception that a hole is designed clay for the process silicon. by Thermalset substituting plastics a metal like epoxy polycarbonates surface coat into the base section of the nose. After fabrication, a locking helicoil is inserted into the hole to provide the can durability be used if required. to make parts for higher strength and threaded surface for attaching to the spindle. The ABS launch lugs, faired inlet, canards, and fins Finally, after all components were fabricated, the Th Aassembly of the model on the strongback proceeded as were all fabricated in the same manner as the fuselage follows: components. The tails and canards were small enough to allow them to be fabricated four at a time in the FDM 0 The forward fuselage section was slid onto the machine. In addition to the rapid prototype ABS fins strongback from the aft end of the strongback until and canards, additional fins and canards were fabricated only the interlocking tab remained. using thefollowingtwo processes: The center fuselage section was mated to the A JHU/APL-developed patent-pending process to forward section via the interlocking tab and the infuse epoxy into the ABS components combined sections slid forward until only the " Casting the components out of resin using a soft center fuselage tab remains off the strongback. mold created from a specially prepared ABS 0 The aft fuselage/boattail section was attached via component the interlocking tab and the three combined The epoxy-infused components are created normally on sections translated forward to align the necessary the FDM machine and surface sanded. They are then holes in the strongback to the matching slots and immersed in an epoxy bath and subjected to a vacuum to remove the air contained within the rapid prototype The nose of the model was attached to the parts and replace it with the resin epoxy. The parts are strongback by screwing the metal spindle into the then removed from the container, excess resin is wiped forward end of the strongback and then screwing from the outer surface of the part, and the resin- the threaded ABS nose onto the spindle. impregnated part is allowed to cure. This results in a FDM rapid prototype plastic part that is no longer At this point the entire fuselage has been completely porous and has significantly greater mechanical assembled and the remaining components added as properties than that of the original part. The final cured needed. For RPM-01, several configuration options part is then sanded and wiped with the Sandfree aerosol were available, with all the optional components solvent to provide the relatively smooth surface used in attached using screws and common screw holes. If the the wind tunnel test. body-alone configuration were to be tested, the filler blocks would be installed in lieu of the canards, inlet, Finally, the third set of fins and canards investigated, and fins. If the full-up configuration was desired, the which showed the most promise for further evaluation, canards are installed using four screws per canard. were the cast resin components. These fins and canards These screws pass through the canard mounting pad were fabricated by first growing a master fin and canard and screw directly into the strongback. The fins and in the FDM machine. These components were then inlet attach in the same way. Because the launch lugs sanded and treated with body filler to fill in surface are not subjected to large aerodynamic forces, it was pores and cracks. Next, a sandable primer paint was satisfactory to screw them into the ABS fuselage using applied to coat the surface. The tail was then staged into helicoil attachments. clay to a break-away plane with care to remove any excess clay from surfaces and edges. A release agent Appendix B - List of References was sprayed on the tail and clay surfaces. A metal frame was constructed around the clay to contain the I. R. R. Heisler, "Final Test Report for the Wind pouring of silicon resin. Tunnel Test of the JHU/APL WTM-01 at the Once cured, the silicon mold was turned over and the Calspan Transonic Wind Tunnel," JHU/APL other half was prepared with release agent and poured Memorandum AlD-1-97U-078, June with silicon to mold the other half of the tail along the UNCLASSIFIED 5-22

14 2. B. E. McGrath and A. T. Hayes, "JHU/APL Model 4. A. M. Springer, K. Cooper, and F. Roberts, Design, Fabrication, and Performance of the "Application of Rapid Prototyping Models to WTMOI Wind Tunnel Model at the Calspan Transonic Wind Tunnel Testing," AIAA Paper 97- Transonic Tunnel," JHU/APL Memorandum AID- 0988, January U-041, June A. M. Springer, "Application of Rapid Prototyping 3. A. M. Springer, "Evaluating Aerodynamic Charac- Methods to High Speed Wind Tunnel Testing," teristics of Wind Tunnel Models Produced by NASA TP , May Rapid Prototype Methods," Journal of Spacecraft and Rockets, Vol. 35, No. 6, November-December UNCLASSIFIED 5-23